Re: Error compiling Glib perl bindings on win32/camelbox

[Brian Manning <elspicyjack gmail com>]

On Thu, Jul 12, 2012 at 4:16 PM, Jeffrey Ray <jeffrey hallock gmail com> wrote:
> After upgrading Glib to to 2.28.8-1 using the binaries from
> http://ftp.gnome.org/pub/gnome/binaries/win32/glib/2.28/ (including the dev
> package).
>
> I attempted compile the perl bindings (version 1.261) I get this output
> error when running perl Makefile.pl
>
> GMainLoop.xs: In function `XS_Glib__IO_add_watch':
> GMainLoop.xs:556: warning: passing arg 1 of `g_io_channel_win32_new_socket'
> make
> s integer from pointer without a cast
>
>
> Is this a bug in the perl bindings?

Depends on how you define the term "bug".  I don't consider compiler
warnings as bugs per se, but that's my opinion.

> Considering it said it was just a warning, ran dmake test anyways, and go
> this output:

[snipped]

> t/a.t ...................... 1/30
> #   Failed test 'default log handler: install undef, prev default'
> #   at t/a.t line 78.
> #          got: undef
> #     expected: 'CODE(0x262e68c)'
> # Looks like you failed 1 test of 30.
> t/a.t ...................... Dubious, test returned 1 (wstat 256, 0x100)
> Failed 1/30 subtests
> t/b.t ...................... ok
> t/c.t ...................... ok
> t/d.t ...................... ok
> t/e.t ...................... 1/312
> #   Failed test at t/e.t line 402.
> #          got: '1'
> #     expected: '2'
>
> #   Failed test 'value 30 value unchanged'
> #   at t/e.t line 404.
> #          got: undef
> #     expected: '30'
>
> #   Failed test 'clamp 70 down to be in range'
> #   at t/e.t line 408.
> #          got: undef
> #     expected: '50'
>
> #   Failed test 'clamp -70 down to be in range'
> #   at t/e.t line 411.
> #          got: undef
> #     expected: '20'
> # Looks like you failed 4 tests of 312.
> t/e.t ...................... Dubious, test returned 4 (wstat 1024, 0x400)
> Failed 4/312 subtests
> t/f.t ...................... ok
> t/filename.t ............... ok
> t/g.t ...................... 1/33
> t/g.t ...................... 1/33
> #   Failed test at t/g.t line 155.
> #          got: undef
> #     expected: 'C:\Users\jhallock\Desktop\Glib-1.261\tmp.ini'
> # Looks like you failed 1 test of 33.
> t/g.t ...................... Dubious, test returned 1 (wstat 256, 0x100)
> Failed 1/33 subtests
> t/h.t ...................... ok
> t/lazy_loader.t ............ ok
> t/make_helper.t ............ ok
> t/options.t ................ ok
> t/signal_emission_hooks.t .. ok
> t/signal_marshal.t ......... ok
> t/signal_query.t ........... ok
> t/tied_definedness.t ....... ok
> t/tied_flags.t ............. ok
> t/tied_set_property.t ...... ok
>
> Test Summary Report
> -------------------
> t/a.t                    (Wstat: 256 Tests: 30 Failed: 1)
>   Failed test:  10
>   Non-zero exit status: 1
> t/e.t                    (Wstat: 1024 Tests: 312 Failed: 4)
>   Failed tests:  303, 305, 307, 309
>   Non-zero exit status: 4
> t/g.t                    (Wstat: 256 Tests: 33 Failed: 1)
>   Failed test:  32
>   Non-zero exit status: 1
> Files=28, Tests=935,  5 wallclock secs ( 0.12 usr +  0.05 sys =  0.17 CPU)
> Result: FAIL
> Failed 3/28 test programs. 6/935 subtests failed.
> dmake:  Error code 255, while making 'test_dynamic'

I don't think any of the test suites for any of the Gtk2-Perl modules
will run cleanly on Windows.  I've always wanted to dig into it and
figure out why, but that hasn't happened yet.

I would keep going with compiling Cairo/Pango/Glib/Gtk2, then try to
run a simple Glib/Gtk2 program as a test.

We would welcome patches against the test suite if you wanted to dig into it.

Thanks,

Brian